Aim: To assess the utility of repetitive nerve stimulation (RNS) in facial and hypothenar muscles in the clinical groups of myasthenia gravis (MG). Patients and Methods: We performed RNS study in the orbicularis oculi (O.O.), nasalis and abductor digiti quinti (ADQ) in 115 consecutive myasthenic patients and classified them according to the classifications of the Myasthenia Gravis Foundation of America. Patients were classified into three groups: group 1, group 2 (IIa, IIIa and IVa) and group 3 (IIb, IIIb and IVb). Results: RNS was abnormal in 95 patients (82.6%): 78.3% in the O.O., 66.1% in the nasalis and 19.1% in the ADQ. Both facial muscles were statistically more sensitive than the ADQ in all groups of patients. RNS in the O.O. was more frequently abnormal than in the nasalis only in group 1. Sensitivity to acetylcholine antibodies in myasthenic patients was 84%. Acetylcholine receptor (AChR) and muscle-specific tyrosine kinase antibodies were present in 96.7% of the patients with abnormal RNS in both facial muscles. Single-fiber electromyogram (SFEMG) was abnormal in 91.3% of the tested patients. One of the three tests used for the diagnosis of MG (AChR antibodies, SFEMG, RNS) was abnormal in 99.1% of the patients. Discussion: O.O. is the most sensitive muscle in all groups of MG followed by nasalis, while the ADQ is the muscle with the lowest sensitivity. Facial muscles, especially the O.O., should be the first to be tested in MG. The negativity of all tests (RNS, AChR antibodies, SFEMG) should question the diagnosis of MG, even in the presence of symptoms consistent with MG.